DE19827704C2 - Method for cylinder-selective knock control of an internal combustion engine - Google Patents

Description

The invention relates to a method for cylinder-selective Knock control of an internal combustion engine according to the generic term of claim 1. Such a method is from DE 29 25 770 A1 known.

When engine knock occurs in a cylinder, the combustion engine, the firing angle for this cylinder is egg delayed a certain amount, which makes the truth likelihood of knocking combustion occurring in the its cylinder is reduced. Do not knock afterwards the engine operation is slowly the ignition angle again by one preset amount adjusted early.

The known total ignition angle for a cylinder in a be The correct operating point is made up of one of load and mo door speed dependent, stored in a map and (every 180 ° KW for a four-cylinder engine) (therefore "cylinder-selective") basic ignition angle GZ (z) at knock-free operation, calculated from that closest to the ignition lying ignition dead center, and from one at this cylinder additional knock adjustment angle due to engine knocking men, whereby it should be noted that the knock adjustment angle only can assume negative signs, according to the definition: a positive sign means a shift to "early", a negative sign, however, a shift to "late".  

The respective knock adjustment angle of each cylinder is ge saves. When changing from one operating point to the next the knock adjustment angle last entered is combined with the new basic ignition assigned to the new operating point angle used.

DE 36 35 963 A1 describes a device for controlling the ignition time of an internal combustion engine is known, in which the Knock adjustment angle depending on the load and speed characteristic map is entered per cylinder. This value when you re-enter this operating point for the Knock control again used as knock adjustment angle.

This has the following disadvantages:

• - is dependent on the time of the operating point change a relatively random knock adjustment angle is saved,
• - there is no exact adaptation of the knock limit,
• - it arises in the case of transitions between the adaptation areas Chen no homogeneous ignition angle curve and therefore none steady torque curve,
• - The internal combustion engine is not exactly at the knock limit operated; this does not guarantee that an op maximum torque and optimal specific fuel sets consumption.

It is an object of the invention, the known method for zy Lindel selective knock control of an internal combustion engine there to improve that as accurate an adaptation the knock limit takes place that at transitions between the adaptation areas a homogeneous ignition angle curve and so that there is also a constant torque curve, and that an optimal torque as well as an optimal speci fuel consumption.  

This object is achieved by a method with solved the features of claim 1.

The invention consists essentially in the fact that to the known cylinder-selective knock adjustment angle KNK (z), a first adaptation value AD1 of a first adaptation circuit and a second adaptation value AD2 of a second adaptation circuit common to all cylinders, the generation of which is explained in more detail below with the aid of a schematic drawing becomes. The total ignition angle ZW (z) for a cylinder z is composed according to the following formula:

ZW (z) = GZ (z) + KNK (z) + AD1 (z) + AD2,

the signs of these values must be observed: GZ (z), AD1 (z) and AD2 can be positive or negative signs have (which can mean early or late adjustment), while KNK (z) only negative sign (late adjustment) ha can.

In the drawing, for a particular one, of load L and Engine speed n dependent operating point of the internal combustion engine seem, the basic ignition angle GZ (z) in relation to the ignition dead center ZOT (z), the course of the cylinder-selective knock adjustment kels KNK (z) in relation to the basic ignition angle GZ (z), and the ver accumulation of the first and second adaptation values AD1 and AD2 carried.

The basic ignition angles GZ (z) are shown in the drawing as an arrow, for each cylinder z, for example, as a crankshaft angle in ° KW, calculated from the ignition dead center ZOT (z) = 0 ° KW of the cylinder z (in the drawing as a horizontal dashed line as a whole plotted below) - early with a positive sign, late with a negative sign - depending on at least engine speed n and load L entered in a map and, as mentioned above, read out and updated from this. With this basic ignition angle, the cylinder z is ignited if no engine knock occurs:

ZW (z) = GZ (z).

However, if engine knock K occurs in cylinder z, the total ignition angle ZW (z) is retarded (decremented) by a knock adjustment angle KNK (z), each time with a step size S _Kdec . After three successive ignitions, knocking strokes K are shown in the drawing, which increase the total ignition angle ZW (z) by a value KNK (z) = 3. S _Kdec on ZW (z) = GZ (z) - KNK (z) = GZ (z) - 3. S _{Shift the Kdec} towards late.

If there is no more engine knock afterwards, the knock adjustment angle KNK (z) is reduced by a predetermined step S _Kinc after one engine cycle (720 ° KW) or several engine cycles (or at predetermined time intervals), enlarged in the drawing in a dashed circle shown, and with it the total ignition angle ZW (z) shifted back to the beginning (incremented).

The step _sizes S _Kdec and / or S _Kinc can be constants or vary depending on the operating point.

In the drawing, two predefined threshold values plotted from the basic ignition angle GZ (z) late, a first threshold value DEC and a second threshold value INC, are entered as dash-dotted lines which are important for the first adaptation value AD1 (z) of the first adaptation circuit :

• - As long as the knock adjustment angle KNK (z) is greater than the first threshold value DEC, t1 to t3 in the drawing, a cylinder-selective first adaptation value AD1 (z) is to be increased with a predetermined step size S _ADdec (whereby the total ignition angle ZW (z) after adjusted late, ie decremented),
• - as long as the knock adjustment angle KNK (z) is smaller than that The first threshold DEC and greater than the second threshold INC is, t3 to t4 in the drawing, intended to be the cylinder  selective first adaptation value AD1 (z) be constant, and
• - As soon as the knock adjustment angle KNK (z) is smaller than the second threshold INC, from t4 in the drawing, the cylinder-selective first adaptation value AD1 (z) _{should be} reduced with a _{predetermined} step size S _ADinc .

In a preferred exemplary embodiment, as long as the knock adjustment angle KNK (z) is greater than the first threshold value DEC, the sum of the knock adjustment angle KNK (z) and the first adaptation value AD1 (dashed line between t1 and t3 in the drawing) is to be incremented with the step size S _kinc who in order not to slow the incremental speed of the total ignition angle ZW (z).

The step size S _Kdec is selected so that the predetermined first threshold DEC, as shown in the drawing, is not exceeded by the knocking adjustment angle KNK (z) at the time t1, for example, but only at the second knocking stroke, z. B. S _Kdec = 3 ° KW and DEC = 3.5 ° KW. This has a dampening effect on the knock control.

The first adaptation values AD1 (z), which are for each cylinder depending on the driving point are stored in a map, and second adaptation values AD2 described below are in Individual diagrams of the drawing are shown on their own, in the diagram above is the sum of GZ (z) + KNK (z) + AD1 (z) shown as a dashed line.

A cylinder-selective takes place in the first adaptation circle Knock adaptation, which depends on the load L and engine speed n the ignition angle differences due to the compression lower adapted between the cylinders. In the map, in which the first adaptation values AD1 (z) are stored, are the cylinder-specific adaptation parts of the knocking  contain containment that determined at the respective operating point were communicated.

The knock limits are usually coordinated with ei a defined engine and defined fuel. Normally engines, components and fuels have certain characteristics Tolerances that significantly affect the knock limit can. These deviations apply to all cylinders; they än change, with the exception of fuel quality only slowly.

A second is used to effectively compensate for such effects Adaptation value AD2 of a second common to all cylinders Adaptionskreis formed in that in the respective Be driving point is the mean AD1 of the first adaptation values AD1 (z) of the first adaptation circle of all cylinders is calculated and is compared with a predetermined threshold S2. Is If the mean value AD1 is greater than S2, AD2 becomes after every Mon gate cycle by a predetermined decrement D (with negative pre characters) in the late direction, otherwise by a predetermined one Increment I (with a positive sign) towards early ver changes. When changing the operating point, the mean value AD1 from the first adaptation values AD1 (z) of the new Be driving point calculated and compared with the threshold S2 and AD2 are enlarged according to the comparison result or downsized.

In the drawing, the bold solid line ZW (z) shows the course of the total ignition angle ZW (z) for a cylinder z from the basic ignition angle GZ (z), knock adjustment angle KNK (z), first adaptation value AD1 (z) and the two common to all cylinders adaptation value AD2:

ZW (z) = GZ (z) - KNK (z) - AD1 (z) - AD2,

where the values KNK (z), AD1 (z) and AD2 each have one Delay the ignition angle. At a time  t2 are the currently valid sizes for ZW (z), KNK (z), AD1 (z) and AD2 shown as arrows.

There is usually no early adjustment of the overall ignition kels ZW (z) beyond the specified basic ignition angle GZ (z) instead of. For example, when changing to a fuel with a higher octane number, however, it can be useful to wear, and also positive values of the first and two adaption, which is an early adjustment of the Total ignition angle beyond the basic ignition angle GZ (z) can interpret.

For physical reasons, however, the operating point areas are determined in which such an early adjustment be permitted beyond the basic ignition angle GZ (z) should and must (possibly operating point dependent) limit values for such an early adjustment to be fixed to engine damage to avoid that.

After each engine cycle (every 720 ° KW) they are currently before values KNK (z), AD1 (z) in the corresponding characteristic diagram depending on the operating point, and the common second Adapti value AD2 is saved and the operating point becomes dependent of load L and engine speed n newly determined.

When the engine is switched off, all first ad AD1 (z) values of the first adaptation circuit in cylinders selective maps and the second adaptation value AD2 of the second adaptation circuit non-volatile (in an EEPROM) saved and read back at the next start of operation.

Due to the separate adaptation _speeds (increments S _ADdec , S _ADinc ) when decrementing or incrementing the adaptation values AD1 of the first adaptation circuit, the mean position of the ignition angle at the knock limit can be determined fairly precisely.

By definition (using the threshold values DEC and INC) of Areas of the knock adjustment angle KNK (z) in which none Adaption is made to solve minor activities of the Knock control no changes in the adaptation values; this leads to the calming of the first adaptation circle and to homogeneous ignition angles.

The knock control component is made up of the knock adjustment angle KNK (z) and the two adaptation values AD1 (z) and AD2 together men. Advantageously, the characteristics of each one NEN circle independent of those of the other circles be put. This makes knock control ideal for everyone Adapt internal combustion engine.

Because the threshold values DEC and INC for activating the first adaptation _{circuit are} linked to the step _sizes S _Kdec , S _{Kinc of} the knock _{adjustment angle} KNK (z), the adaptation characteristics are not influenced when the knock control _parameters are changed.

The input variables for the second adaptation circuit are first adaptation values AD1 (z) of the first adaptation circuit, where can be exploited by its dampening effect; it there are only ignition angle corrections in the second adaptations circle caused by aging effects or changes in strength fabric quality.

Claims (6)

1.Method for cylinder-selective knock control of an internal combustion engine, at which, assigned to the respective operating point dependent on load (L) and engine speed (n), a predetermined, non-knocking operation assigned, cylinder-selective basic ignition angle (GZ (z)) and one for each Knocking (K) occurring gradually increase in the direction of "late" with a predetermined step size (S _Kdec ), and if operation is not knocking after each engine cycle (720 ° KW) stepwise in the direction of "early" with a predetermined step size (S _Kinc ) decreasing, cylinder-selective knock adjustment angle (KNK (z)) form a cylinder-selective ignition angle, characterized in that
that a cylinder-selective first adaptation value (AD1 (z)) of a first adaptation circuit is determined from a comparison of the knock adjustment angle (KNK (z)) with a first (DEC) and a second threshold value (INC),
that the first threshold value (DEC) is greater than the step width (S _Kdec ) of the cylinder-selective knock adjustment angle (KNK (z)) which is predetermined in the "late" direction,
that the second threshold value (INC) is smaller than the step width (S _Kdec ) of the cylinder-selective knock adjustment angle (KNK (z)) which is predetermined in the "late" direction,
that the cylinder-selective adaptation value (AD1 (z)) of the first adaptation circuit with a predetermined first adaptation increment (S _ADdec ) after each engine cycle (720 ° KW)

• 1. is changed in the "late" direction as long as the knocking adjustment angle (KNK (z)) is greater in amount than the first threshold value (DEC),
• 2. is kept constant as long as the knock adjustment angle (KNK (z)) is smaller than the first threshold value (DEC) and larger than the second threshold value (INC), and
• 3. is changed in the "early" direction with a predetermined second adaptation _{step size} (S _ADinc ), as long as the knock _{adjustment angle} (KNK (z)) is smaller in amount than the second threshold value (INC),

that a second adaptation value (AD2) of a second adaptation circuit assigned to all cylinders (z) is determined from a comparison of the mean value (AD1) of all cylinder-selective adaptation values (AD1 (z)) of the current operating point with a predetermined threshold (S2), and
that the cylinder-selective first adaptation value (AD1 (z)) and the second adaptation value (AD2) assigned to all cylinders (z) to form a cylinder-selective total ignition angle (ZW (z)) according to the formula
ZW (z) = GZ (z) + KNK (z) + AD1 (z) + AD2
can be used, with z = cylinder number. 2. The method according to claim 1, characterized in that the second adaptation value (AD2) of the second adaptation circuit

• 1. After each engine cycle, a predetermined decrement (-D) is changed in the late direction if the mean value (AD1) of all cylinder-selective adaptation values (AD1 (z)) of the current operating point has a negative sign and is larger than the threshold (S2) is and
• 2. After each motor cycle, a predetermined increment (+ I) is changed in the early direction if the mean value (AD1) is smaller than the threshold (S2).

3. The method according to claim 1 or 2, characterized in that at least one of the sizes first step size (S _Kdec ), second step size (S _Kinc ),
first adaptation step _size (S _ADdec ), second adaptation step _size (S _ADinc ), first threshold value (DEC), second threshold value (INC), is stored in a map assigned to this variable depending on the operating point. 4. The method according to any one of claims 1 to 3, characterized ge indicates that predetermined operating point ranges for a Adaptation of the first or second adaptation circuit blocked are. 5. The method according to any one of claims 1 to 4, characterized in that for the adaptation values (AD1 (z)) or (AD2) for all cylinders (z) valid maximum values (+ AD1 _max , + AD2 _max ) in the early or Minimum values (-AD1 _min , -AD2 _min ) are specified late. 6. The method according to any one of claims 1 to 5, characterized in that during non-knocking operation in a loading area (t1 to t3) in which the knocking adjustment angle (KNK (z)) is greater than the first threshold value (DEC), the sum of the knock adjustment angle (KNK (z)) and the first adaptation value (AD1 (z)) is incremented with the step size (S _Kinc ).